From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 1 Aug 2002 08:20:25 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 1 Aug 2002 08:20:24 -0400 Received: from pc2-cwma1-5-cust12.swa.cable.ntl.com ([80.5.121.12]:54510 "EHLO irongate.swansea.linux.org.uk") by vger.kernel.org with ESMTP id ; Thu, 1 Aug 2002 08:19:42 -0400 Subject: Re: 2.4.19-rc5 from 2.4.18 From: Alan Cox To: Bruce Cc: linux-kernel@vger.kernel.org In-Reply-To: References: Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.3 (1.0.3-6) Date: 01 Aug 2002 14:39:30 +0100 Message-Id: <1028209170.14871.1.camel@irongate.swansea.linux.org.uk> Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 2002-08-01 at 11:21, Bruce wrote: > I have a problem that has manifested itself since 2.4.19-rcx. > > An ADSL driver (1483-2.4.16.o) for the DLINK dsl100d PCI ADSL modem > operational under 2.4.18 has been broken since 2.4.19, > specifically under rc3,rc4 and rc5. (Haven't tested rc2 or rc1) > > It appears to get/interpret an incorrect hardware address from the card. > > There is a constraint. The driver was originally compiled for ITEX/DLINK > under 2.4.16 and I don't have access to the source. You will need the source to deal with it. Binary only modules have to match the exact kernel properties, so you effectively threw away all the freedom you got from Linux by relying on it.